"Tuning" is a term used freely in the automotive aftermarket. Perhaps a bit too freely. Put plainly, all of the tuning we're discussed in the past wasn't real tuning. Instead, it was an exercise in trial and error. The process was simple: We would press a button a bunch of times until we found our desired outcome, whether it be a certain boost level or a specific oxygen sensor voltage. We would then continue to make semi-educated assumptions based on incomplete data. We had no clue as to what ignition advance we were running at full boost. Injector duty cycle? Detonation threshold? Knock sensor authority range? Who knows! A couple of these lapses of knowledge with much-higher-than-stock cylinder pressures, and we were heading for trouble. Understandably, in such a situation, we chose to be grossly conservative with our engine tuning. But "conservative" does not mean "good." With overly conservative ignition timing, our exhaust gas temperatures would rise to uncomfortable levels. With too much fuel, our engine would misfire, backfire dramatically, lose power and carbonize its pistons. To put it quite simply, our Project Impreza had become the automotive equivalent of Frankenstein's monster. Strong, unpredictable, dumb, somewhat uncontrollable and frighteningly self-destructive. Fortunately, our recent installation of the Electromotive TEC-II engine management system should change all that.
"Total Engine Control" isn't just a catchy term, it's a necessity when trying to double the horsepower output of any engine--let alone one with a relatively high 9.7:1 compression ratio running more than 10 pounds of boost on California's pathetic 92-octane squirrel pee. Being able to create custom ignition and fuel maps should enable us to maximize drivability and performance while achieving a margin of safety one wouldn't expect from such a highly stressed powerplant. Last month we detailed the installation of the TEC-II--that was the easy part. Now we must tune it.
Let the tuning begin!
With our laptop connected to the TEC-II via a long serial cable (which was routed through a grommet in the firewall), the first step is to fire up the Electromotive WinTEC software and conduct our diagnostic sensor check to see if we wired things correctly.
With all of our sensors (throttle position, MAP, coolant temperature, intake temperature, etc.) appearing to be fully functional, we can now create our tuning maps, or calibration file.
Building a calibration file
Fortunately, Electromotive includes a baseline calibration file in every software bundle. Unfortunately, we're more likely to see a purple cow stuck in our chimney than to actually have it work well on our turbocharged Subaru. This means we actually have to use our brains and create our own file. And furthermore, that also means we have to understand what the heck is going on and, more importantly, how the TEC-II thinks.
First, it's obvious that the TEC-II doesn't work like other engine management systems. A cursory glance through the software reveals no vast fuel tables full of countless injector pulse-widths or duty cycles. Instead, we see a small eight-by-eight grid known as the Volumetric Efficiency Table. Hardly impressive looking, compared with other engine management systems that boast far more rows and columns.
But according to Electromotive, the TEC-II approaches the fuel delivery game from a completely different angle. Its unique approach is based upon something called the Theory of Linear Thermodynamics. What the heck does that mean?
Quite simply, it assumes that an engine, which is basically an air pump, will fill each cylinder with a given amount of air under a given manifold pressure. So, at 2 bar of absolute manifold pressure, the engine will ingest twice as much air as it would at 1 bar of absolute manifold pressure. Let's take our favorite turbocharged EJ25 as an example. For argument's sake, let's say that we want to run the engine at a maximum of 2 bar of absolute manifold pressure (that's 1 bar of boost on top of 1 bar of atmospheric pressure). Under this full load condition, let's assume that in order to achieve a stoichiometric 14.7:1 air/fuel ratio, we need to keep our fuel injectors open for X milliseconds for every single combustion event. With this simple relationship, we can rest assured that if we were to hold everything else constant, but reduce the absolute manifold pressure (also known as "engine load") to 1 Bar, our injector on-time falls to just 0.5X milliseconds. In fact, using simple fifth grade algebra, we could theoretically map the fuel requirements for our engine under any conceivable condition by just solving for X! Something a computer can do at breakneck speeds without ever missing a beat.
But what about engine speed? Doesn't that affect injector on-time as well?
In theory, no, it doesn't. In fact, injector on-time is dependent on engine load and nothing else. Engine speed, for the most part, only defines how much time there is between each engine event. As engine speed increases, the length of time between each engine event decreases. This means that we will eventually run into a bottleneck in which the time period between each engine event equals the desired injector on-time. When this happens, our injectors are operating at 100-percent duty cycle. But more on that later.
But what about the volumetric efficiency characteristics of each engine? That doesn't remain constant under all engine conditions!
Absolutely! And that's exactly what the aforementioned Volumetric Efficiency Table is used for. This eight-by-eight table is used to "massage" the raw fuel curve (which, as we've seen, is really a line defined by X) to match the pumping efficiencies of any given engine, be it a slow-spinning diesel to a high-revving 1.8 VTEC. In other words, as other engine management systems compute fuel demands by constantly referring to look-up tables and interpolating between adjacent cells, the TEC-II constantly runs the simple fuel algorithm:
Say what!?!
While this algorithm may sound unintelligible, let's break it down to its individual components.
·MAP% is nothing but engine load as measured by our MAP sensor.
·TOG is the abbreviation for Time On Gamma wish is just a fancy term for the coefficient what we've been describing as "X." This value, like "X," is measured in milliseconds.
·GAMMA is nothing more than a Greek-named multiplier that is dependent, among other things, by the values in the Volumetric Efficiency Table. With a GAMMA of 1.0, fuel is delivered according to the Theory of Linear Thermodynamics. With a GAMMA of 0.9, 90 percent of that fuel is delivered. Likewise, with a GAMMA of 1.10, 110% of that fuel is delivered. The reason we need such a multiplier is because, as we know, a real-life engine is not 100% thermodynamically linear. In other words, its volumetric efficiency changes slightly with load, engine speed and desired air/fuel ratio--and so must GAMMA.
·IOT is Injector Offset Time, which could be positive or negative, and is used primarily to, you guessed it, offset the entire raw fuel line up or down. This value, is primarily used to achieve stable injector on-times at idle. Go back to fifth grade algebra and think of GAMMA as the slope of a linear equation and IOT as the Y-intercept. Don't worry, it sounds more complicated than it really is.
Getting started
Now that we understand the TEC-II's operating fundamentals, let's see if we can get our engine started. Using the WinTEC Tuning Wizard, we can create our first calibration file. Not only does this feature get us off to a good start; it also tells us if our fuel system is up to the task of supporting the horsepower numbers we are aiming for. According to the Tuning Wizard, with 550cc injectors (at a standard 43-psi fuel pressure) we should be able to safely support 320 hp, assuming a Brake Specific Fuel Consumption (BSFC) of .60. This jibes well with the injector math prescribed at RC Engineering that we discussed last month.
With our calibration file now created and saved under the arbitrary name, "test-1," we are brought to WinTEC's main menu. It is in this window that we can open, edit, save and close files. This is also where we will go to download files into the TEC-II, as well as creating and viewing datalogs. Obviously, Tuning Wizard or not, we still have a good deal of tuning to do. To begin tuning, we click on "Edit a Calibration File" and select "test-1." This takes us to the Calibration Screen which will be used to program every aspect of engine management, including the ignition table, the air/fuel ratio table, VE table, rev limits, boost maps, fuel enrichments, knock sensor control, warning lights and a few dozen other things. In other words, we are going to be spending a fair amount of time in this window, jumping from parameter to parameter.
Our first stop will be in the Basic Injector Fuel Graph window where we will set our minimum injector turn-on time to .9ms, which, according to RC Engineering, is the shortest on-time a low-resistance injector can achieve without stalling.
Our next stop is the Ignition Advance Table where we will create our baseline ignition map. With eight user-definable MAP and Engine Speed breakpoints, we can tailor an ignition map to best suit our needs. Right off the bat, we noticed that the rpm and MAP breakpoints are equidistant from each other. While this looks nice and neat, it doesn't take full advantage of the TEC-II capabilities. Our first rpm breakpoint will stay at 1000 rpm. This is to ensure a smooth transition from idle speed to ultra low lugging speeds. We'll set our next breakpoint to 2500 rpm, which seems like a nice, low, around-the-town cruising speed. The next breakpoint will be 3500, which is the engine speed at which our turbo motor makes maximum boost. For the remaining breakpoints, we pick 4200, 5000, 5500, 6000 and 6500.
Ideally, you want to set your breakpoints to the engine speeds at which torque output changes, because when torque output changes, fuel and ignition demands change as well. When torque output is "flat" between two rpm points, fuel and ignition requirements remain, for the most part, static. But with no real dyno graphs to reference off of, we're just playing it by ear.
Next up for tweaking is the MAP breakpoints. Just to make your life more confusing, MAP values are in kilopascals, or kPA. 100-kPA is the same as 1 bar, or 14.7 psi, or atmospheric pressure. We'll set the lowest MAP breakpoint to 35-kPA, which is the approximate load value we would see under engine braking. We'll set the next breakpoint for a typical cruising load of 65-kPA. Since the transition from vacuum to boost operation often necessitates an immediate reduction in ignition advance and a sudden change in desired air/fuel ratio, we'll set the next breakpoint at 100-kPA, which is atmospheric pressure. With the first three established, we have five breakpoints remaining to cover loads from 100-kPa to 200-kPa (0-psi to 14.7-psi). With a high compression engine and relatively low available octane, we suspect that tuning is going to become a little tricky at loads greater than 140-kPa. That being the case, we'll spread the remaining breakpoints evenly from 140-kPa to 200-kPA. Between the breakpoints, the TEC-II makes up new values based on a straight line between those values. In other words, if a breakpoint at 5000 rpm has a value of 10 degrees, 5500 rpm is at 15 degrees, the TEC-II will use 12.5 degrees at 5250 rpm.
With 30 degrees of total ignition advance at 200-kPa (that's 14.7-psi of boost!), experience tells us that the current map is far too aggressive for our high compression Subaru motor. For now, we set the highest MAP row (200-kPa), between 3500 and 6500 rpm, to a very conservative 10 degrees of advance. Below 3500, we'll reduce the ignition advance even further, since low engine speeds, all other things being equal, require less ignition advance. It should be noted that we'll never actually operate our engine in this region of MAP, since we don't expect such boost pressures until well after 3000 rpm. On the other end of the load axis, at 35-kPa between 3500 and 6500 rpm, we'll set the advance to 40 degrees.
For now, we'll simply "smoothen out" the rest of the cells by mentally linearizing between our previously defined rows. With all of our ignition cells defined, we'll set our Initial Advance (which covers all the ignition values below our lowest 1000 rpm breakpoint) to 21 degrees. To step back and take a look at the whole map, we'll view the 3D graph and check for strange outliners or rough spots. With none seen, we'll move on to the Air/Fuel Ratio Table.
Before we start adjusting values, let's separate fact from fiction from hearsay. All of us have heard tuners refer to specific nominal air/fuel ratios which are best for power. Some say it's 12:1. Others may argue that it's 12.5:1. Whatever. The truth is that these nominal values mean next to nothing without the ability to actually enter the combustion cylinder and count how many air molecules there are for each fuel molecule. Without this ability, these desired air/fuel ratio is arbitrary guess. The fact is, that most tuners derive these nominal values from 02 sensor voltages. But since most tuners reference these voltages from a narrow band oxygen sensor (which, for the most part, is only accurate for measuring near-stoichiometric air/fuel mixtures), their findings are often clouded in shadowy doubt. So, for now, we're going to forget about everything we've been told about air/fuel ratios and play it by ear. And our ear, based on previous experience with Electromotive engine management systems, tells us that 13.5:1 is about as rich as we would ever want to run under boost. With that in mind, we'll temporarily fill the top three rows with 13.5 values. The next two rows will run slightly leaner (but still quite rich) at 13.6 and 13.65, respectively. With the bottom three rows representing off-boost performance, we can afford to be a little more aggressive with our air/fuel ratios. In fact, in order to maximize fuel economy under highway cruise conditions, we'll shoot for relatively lean 15:1 air/fuel ratios between 2500 and 3500 rpm. But at higher rpm, under low load, we'll going to run slightly rich in order to facilitate engine cooling. Looking at the 3D Air/Fuel Ratio graph, we can review and check our work.
Next, we'll go to the EGO (Exhaust Gas Oxygen) Parameter and configure the TEC-II's O2 sensor feedback logic. Typically, O2 feedback systems allow for 10 to 15 percent closed-loop correction under low loads to optimize fuel economy and minimize emissions. Under heavy loads and high engine speeds, however, this feedback system typically turns off, making the ECU "fall back" on a set of pre-determined fuel maps. This represents the transition from closed-loop to open-loop behavior. With the TEC-II, we can configure our 02 feedback system a number of ways. Not only can we determine how much authority range to give it, we can also determine under what conditions (if any) the TEC-II transitions from closed-loop to open-loop behavior.
First, let's establish the strengths and weakness of the closed-loop feedback system. At first glance, one may ask, "Heck, it's constantly adjusting fuel delivery to achieve a desired air/fuel ratio. Why disable it at all? The system will never go out of tune, right?" In theory, he would be correct. In application, however, there is more than meets the eye.
The sad fact of tuning is that engine sensors--oxygen sensors included--don't always offer the most accurate information. Nor do they always react quickly enough. Such is the case in our situation when even the slightest amount of sensor lag can result in dangerous lean spots, potentially inducing piston-obliterating detonation. And it's just not the sensor lag that complicates things. In fact, the feedback system itself, by nature, tends to obscure and blur transients by reading a finite number of samples in a given period of time and averaging them periodically. This takes time. Also adding to the confusion is a feedback system that, by nature, is never quite happy with what it reads. Instead, it always tries to correct and correct and correct, rarely reaching any homeostasis.
Fortunately, the TEC-II gives us the ability to fine-tune these aspects of feedback systems. Not only can we define the number of samples between averages, we can also define how large or small the fuel correction will be. But we have to be aware of the consequences: While increasing the number of samples and reducing the correction amount will yield a very stable, non-oscillating system, it will also result in a system that is slow to respond. The converse holds true as well: With a reduced number of samples and a large correction value, the system, while able to respond quickly to changes in air/fuel ratios, may become quite unstable and prone to oscillation. For baseline tuning purposes, we'll go with the default values, only increasing EGO Authority Range from 15 percent to 40 percent. With the EGO deactivation switches set to 7000 rpm and 208-kPa, the feedback system will remain on at all times. This is absolutely necessary, as we will be relying, for now at least, on the EGO feedback system to achieve our desired air/fuel ratios.
And now... the moment of truth
With our baseline calibration file modified, saved under yet another name and downloaded into the TEC-II, we can actually start the engine. And start it does. On the first attempt in fact. With the Engine Monitor Screen displayed on our laptop, we keep a close eye on things. Moments later, the unpleasant aroma of unburnt fuel and the rich air/fuel ratios suggest we will eventually have to adjust our cold start enrichments. But once up to temperature, Project Impreza burbled happily, idling at a slightly bouncy, but otherwise normal, 750 rpm.
With the car idling somewhat acceptably and running a completely untested Time On GAMMA (TOG) and Injector Offset Time (IOT), the next step was to dial-in the appropriate raw fuel curve. This required actual driving and real-time engine monitoring (or datalogging). To set the correct raw fuel curve we had to load test the engine, under wide-open throttle, from low rpm, all the way to redline. Our goal was to find the TOG and IOT values that resulted in the least EGO correction at the peak horsepower point.
Say what? Since we temporarily gave the TEC-II authority to used closed-loop control all the time, it will automatically find its way to the right air/fuel ratio. But in doing so, it will use an EGO (that's Exhaust Gas Oxygen, or O2 sensor) correction to deviate from the rough raw fuel curve we built. If we tweak the TOG and IOT values to make this deviation as small as possible, we will end up with a much more accurate raw fuel curve.
Since the accuracy of our fuel curve is most important at the power peak, which we will assume, for now, is 6000 rpm, we want the EGO correction to be the lowest there. And with an untested ignition map and default fuel enrichment parameters, we had to keep our ears open for detonation, re-tuning our ignition table accordingly, if detected. After several wide-open throttle test runs and no signs of detonation, we finally arrived at TOG that best matched our fuel requirements, with minimum EGO correction, at the peak horsepower point. With our TOG value set to 10.5 ms, we could now tune for IOT, which is used primarily to stabilize idle conditions.
With an IOT of 0 and TOG of 10.5, we noticed a heavily negative EGO correction in our Monitor Screen at idle. In order to tune out the correction, we began to lower IOT, in 0.125 ms increments, until correction oscillated just around 0. With IOT set to -0.625 and TOG set to 10.5, our raw fuel curve was complete! But our idle still remained unstable. We also noticed our car almost stalls on deceleration, right off idle. Looking at the EGO correction during these conditions, we can see why: It's all over the place! Looking at the injector on-time, we see that it is varying between 1.0 and 1.3ms, with the lower range being just beyond the capabilities of the injector. If only there was a way to define a minimum injector on-time that prevented the injector from firing below a certain duration.
Fortunately, there is. Jumping back to the Basic Injector Parameter screen, we change the minimum injector on time from 0 to a far more reasonable 1.15ms. Downloading the change and starting the car again changed the situation. For now, the EGO correction would begin a downward spiral, eventually pinning itself at negative 40 percent. And once the throttle was touched, the car would stall immediately. Apparently, with our higher minimum injector on-time and 14.3:1 air/fuel target, we were asking the TEC-II to do something it couldn't do. To fix this, we slowly began to reduce our minimum injector on-time until this problem disappeared, eventually settling on 1.05ms. This finally solved the stalling annoyance, only to replace it with an oscillation problem. Looking at the engine monitor screen, we noticed the engine would oscillate, hand in hand, with EGO correction. To fix this, we went back into the EGO Parameter screen and adjusted our EGO Reactivity, or the size and speed of EGO correction. Slowing things down, we finally achieved perfect idle quality.
As you can see, achieving a consistent, smooth idle is one of the biggest challenges with stand-alone engine management systems. It is not at all uncommon to see otherwise well-tuned cars that stall or suffer from off-idle bogs. But as involved as our idle quest was, keep in mind that we were asking the car to idle with fuel injectors that are twice as big as the ones it came with. Not bad, eh?
Filling in the VE table
With the raw fuel curve properly defined and the ignition map looking halfway decent, Project Impreza already drives surprisingly well. But leaving things as they are now would be a pity and a waste, given the TEC-II's extraordinary tunability. Now, it's time to fill in the VE table. Properly set up to match the volumetric efficiency characteristics of our turbo EJ25, the VE table will ease the burden off our EGO feedback system which, under certain situations, is correcting plus or minus 15 percent. For example, according to our data-logs, at 65-kPA of load at 3000 rpm, EGO correction--or the amount of correction necessary to achieve the desired air/fuel ratio--oscillates between -10 and 0 percent. No doubt, this is due to our lean air/fuel targets during cruise conditions. To bring our correction values down close to zero, we'll insert a "-5" in the 3500 rpm/65-kPA cell.
As another example, at 5000 rpm and 155-kPa, we see a EGO correction of 2 to 6 percent, suggesting that our engine needs more fuel to achieve the desired rich air/fuel ratio targets. In this situation, we'll insert a "6" in the appropriate cell. (While we could insert a "4" into the cell, a "6" would provide the safest performance by essentially adding a temporary enrichment until the EGO feedback system makes its final adjustments). Likewise, at 6500 rpm and 155-kPA, EGO correction is a whopping -7 to -10 percent, suggesting that our engine, like most engines, is losing pumping efficiency at higher engine speeds. Again, we'll reduce EGO correction by tweaking our VE table accordingly. After a few hours of datalogging and fine-tuning, our VE Table was finally complete.
As fascinating as VE table tweaking may be, by the time you read this story, Electromotive will have released WinTEC Verson II. This new software package will offer, among other things, a new "auto calibration" feature. Using software-based trickery, the TEC-II will be capable of filling the VE table all by itself through the analysis of datalogs. The entire process would involve driving around for 10 to 20 minutes, with the TEC-IIs datalogger downloading directly to the hard drive on the laptop computer. Once finished, the datalog file would be analyzed by the "auto-calibrator," which would then dial-in the VE table accordingly. Tuning really doesn't get much easier than that, does it?
Filling in the ignition advance table
If there was ever a time to proceed carefully and conservatively, this is it. Ignition advance, by far, is the most critical component of any high-output engine management profile. With 2 to 3 degrees, at times, being the difference between safe operation and piston-mulching detonation, one needs to take one little step at a time. The truth is, a poor ignition advance table can kill any engine. Not just from too much advance, either. Too little ignition advance, which results in dangerously high exhaust gas temperatures, can cause big problems as well.
For this reason, we are going to use to use two of the most helpful ignition tuning devices available: an EGT gauge and a good set of ears. With the TEC-II's programmable knock sensor disabled, we can slowly start advancing our timing (through the Add/Subtract Advance feature in our Engine Monitor Screen) until we hear mild detonation. We do just this for as many rpm/MAP combinations as possible, adjusting our ignition table accordingly. Again, the importance of taking small steps cannot be overstated. It is also important to immediately lift off the throttle once detonation is detected. While short periods of mild denotation won't destroy an engine, prolonged periods most definitely will. Many hours and several downloads later, we had our ignition advance map complete. With absolutely no signs of detonation, exhaust gas temperatures below 1550 degrees F, or low EGO corrections, we could now move on to the next step in the tuning process.
Turnin' up the boost
Up until now, we've been running without any help from the TEC-II's electronic boost control feature. With our boost levels set to a mechanically limited 6- to 7-psi of boost (around 145- to 150-kPa of absolute pressure), Project Impreza was running very quickly and very safely. But with boost pressure dropping above 5000 rpm, our softly sprung wastegate needed some electronic assistance. Using the TEC-II's General Purpose Output (GPO), we can tune our boost control solenoid to manipulate the wastegate signal, allowing us to achieve almost any combination of rpm-dependent boost level and spool-up rate. The bigger the numbers we input the GPO table, the greater the solenoid's duty cycle, which means, in turn, the greater the boost-level. Starting off with small numbers and working our way upwards, we gradually start bumping up boost levels until we reach a desired peak boost pressure of 170-kPa (about 10-psi). To eliminate the top-end boost roll-off, we increase the duty cycle in the appropriate cells. It's as simple as that. The final results are a perfectly consistent boost curve with absolutely no transient spiking, oscillating or creeping. The fanciest of aftermarket boost controllers could do no better.
Odds and ends
They say that God is in the little details. In the case of engine management, they couldn't be more accurate. In addition to dialing-in the raw fuel curve, VE table, ignition advance table and GPO table, we still have a good deal of work to do before we can consider our tuning complete. But instead going through each and every additional tuning feature, explaining what we did and why we did it, we're just going to pick a few of the more important functions and explain their importance.
Fuel enrichment parameters
The TEC-II has a bunch of them. When it comes to drivability, fuel enrichments can make or break a car's subjective "feel." If one is too generous with fuel enrichments, the car can feel "boggy" during sudden changes of load or throttle positions. If one is too stingy with fuel enrichments, the engine can temporarily lean out, causing detonation, misfire and a general feeling of "yuckiness."
Fortunately for us, the TEC-II comes armed to the teeth with many a multitude of fuel enrichment parameters, each employed and configured for a different situation. Used primarily for acceleration enrichments, the MAP and TPS triggered enrichments offer user-definable sensitivity, duration and enrichment amount. On the other end of the enrichment spectrum is the Deceleration Fuel-Cut parameter, which is used to determine, among other things, when and how the fuel is cut during coast down. Also vital to for maximum and safe engine per- formance use is the MAT (Manifold Air Temperature)/Voltage Compensation parameter.
With this feature, the TEC-II can adjust fuel and ignition as a function of intake temperatures. Due to exhaust gas temperature limitations, we can't afford to be too conservative with our timing map. This feature allows us to run more advance when we can, only retarding ignition when intake temperatures rise. When three degrees of unnecessary ignition retard can raise exhaust gas temperatures by 150 degrees C, this feature is extremely useful. Another novel feature, but one we currently have no use for, is the MAP/TPS (Throttle Position Sensor) Blend feature will allow us to use an ultra-aggressive high overlap cam profile, yet still tune for stable idle and low-load characteristics. Finally, we come to the self-explanatory Starting Enrichments and Warm-up Enrichments, both of which are used to ensure proper fuel delivery under, you guess it, start-ups and warm-ups!
Idle speed control
For our Parallel Engine Management application, this feature is disabled. We are, after all, still relying on the stock ECU and idle control motor to maintain a nice and steady idle. The less tuning, the better.
Sensor parameters
This feature allows us to define the failure voltages and default setting of all our sensors. As a nice added benefit, it can also be used to trigger the stock Check Engine light when coolant temperatures, boost pressures or intake temperatures go beyond a certain, user-definable limit. If MAP pressures become too high, the default MAP value can be set low enough to create an engine-saving fuel cut.
Knock control
Enabling and programming the knock sensor feature is best left until after the rest of the tuning is complete. This way, one doesn't rely on the knock sensor for tuning. Instead, the knock sensor becomes an active safeguard, ready to jump in when the need arises. With programmable knock thresholds, adjustable rate of retard and advance, maximum allowed ignition retard and rpm-triggered inhibition, the knock sensor system can be tailored to match virtually any engine characteristic. Once tuned, it will provide an added margin of safety against a bad tank of gasoline, excessive heat-soak or any other unforeseen and harsh situation. A big bonus.
Tuned and on the road...
Going from a flurry of feature-limited, low resolution piggy-back computers to a powerful, well-integrated, stand-alone engine management system was sure to yield some extra power and drivability. But it wasn't until we experienced it ourselves, did we realize the extent of the improvement. Saying our project car behaved like an OEM turbo car wouldn't be accurate, simply because very few OEM turbo cars are this absurdly fast.
First gear rips by in a blink, with the tires actually loosing a little bit of traction just before the shift into second gear! Second gear doesn't take much more time either. In fact, it's not until third gear that we have enough time to relax and think about how quickly we got to 80 mph. More impressive than the raw power, however, is the engine's current state of well-being. Simply put, detonation is not a problem. In fact, despite self-induced, extended periods of idling (and the resultant intercooler heat-soak), we have yet to hear a single ping. Even sizeable changes in atmospheric conditions, from the heat of blistering summer day heat to the cold, chilly nights near the coast, the weather has yet to throw a wrench in our highly adaptive engine management system.
With a 2 to 3 degree variance in ignition advance capable of changing exhaust gas temperatures by 150-200 degrees F, precise and consistent ignition timing is critical in any application--especially when turbocharging a relatively high compression engine.
A nice feature of the WinTEC software is the ability to view and manipulate the ignition, VE and air/fuel ratio tables in 3D graph, as well as table format. Viewed in table format, simply move the mouse to the appropriate cell, punch in the value with the keyboard and hit "Enter." Viewed in 3D Graph format, toggle through the map using the arrow keys. To adjust the values, use the plus and minus keys. The ability to look at the "shape" of all the graphs allows one to notice trends and outliners that would otherwise go unnoticed.
Looking at our 3D Air/Fuel Ratio Graph, we can see the leaner air/fuel ratio targets in the mid-band during light load conditions. This, in addition to aggressive ignition timing values, is the secret to getting better-than-stock fuel economy during highway cruise conditions. Last time we checked, we logged 31 miles-per-gallon during a Los Angeles to San Francisco journey.
When it comes to closed-loop fuel management, this table is the ultimate authority. Through the closed loop EGO feedback system, TEC-II with adjust injector pulsewidth in order to achieve the values in this table. Fortunately, proper tuning of the VE table and EGO parameters keeps fuel delivery more proactive and less reactive.
It's hard to believe that plugging in just two numerical values will create a functional raw fuel curve for our turbocharged EJ25. Certainly a lot easier than filling out more than a hundred cells with injector pulse widths.
Herein lies the beauty of laptop programmability. The ability to jump from parameter to parameter with the click of a mouse translates to fast and efficient tuning. Anyone with any Windows experience will find WinTEC easy to navigate through.
With this table, it is possible to create some surprisingly intricate boost patterns. It's possible to program a hefty overboost in the midrange, with boost falling off gradually at the top end as injectors reach maximum safe duty cycle. But for now, we're going to tune for a nice, steady 10 psi across the board.
In case you didn't know, the fuel injection system in your car is programmed to turn off the injectors during engine coasting. This is the main cause of the "jerk" you feel when you lift off the throttle. Jerky or not, doing so helps fuel economy.
The Volumetric Efficiency Table is the key to quick, immediate and accurate fuel delivery. Without it, one would have to rely on the relatively slow EGO feedback system to provide proper fueling. Properly tuned, EGO correction should never exceed a few percent.
Extra fuel is necessary to fire up an engine--even a warm one. Without this short, but generous, enrichment, the engine would have to be turned over several times before it comes alive.
The ability to change fuel and ignition advance values as a function of intake temperatures allows us to tune more aggressively.
This enrichment, which is triggered by sudden changes of manifold pressure (MAP), is used primarily in turbocharged cars, to prevent transient lean-run conditions under sudden bursts of acceleration.
A good, active knock sensor system is absolutely vital for high output applications. There are far too many factors beyond our control to rely completely on our ignition maps. Although Electromotive suggests disabling the knock sensor above 4500 rpm, due to high-speed engine noise, which can be misinterpreted as detonation, we found that with the silky smooth EJ25, we could reliably keep it up and running all the way to our slightly higher-than-stock 6500rpm fuel cut.
To customize the rpm and MAP breakpoints for your particular engine, we go to the Ignition Advance Table. If we were tuning a heavily cammed race engine, which was designed to operate in the 6000 to 9000 rpm range, we would set the breakpoints accordingly.
One look at a properly tuned VE Graph can tell you the precise breathing characteristics of any given engine. If an engine needs more or less fuel to achieve the desired air/fuel ratios, one can also assumes it is consuming more or less air mass.
WinTEC not only controls idle speed through the manipulation of an idle control stepper motor, but also through small ignition timing changes.
In a dynamic system such as an engine, it's necessary to be able to fully tune fuel enrichments. More than anything else, fuel enrichments contribute to drivability and subject "feel." Tuned improperly, no amount of Basic Injector Parameter, Air/Fuel Ratio table and VE table tweaking, will keep the engine running well.
The engine monitor screen provides a real-time display for every engine parameter, from air/fuel ratio, to manifold pressure, to injector duty cycle, to engine speed, to EGO correction, to... er... you get the idea. As a thoughtful touch, WinTEC offers a "high contrast" screen option which makes viewing easier when programming in direct sunlight. Your eyes never had it so easy!
Want to know why the WinTEC software takes up so much hard drive space? Help menus. Lots of them. One menu for pretty much every engine parameter. Just hit the "F1" key and prepare to read. Sometimes a little more than you would want to...
Firing up the WinTEC software brings us to this screen, which is used to swap between files, to monitor the engine, access cute little tuning aids (like a calculator and notepad), and to record and review graphical data logs.
Of all the parameters, this one takes the longest to tune because doing so involves tuning on a cold engine. But once complete, the car should behave just as obediently as it did when it was stock.
A little help never hurt anyone. That's the idea behind WinTEC's Tuning Wizard feature, which helps create the basic "skeleton" of a new calibration file. While it doesn't do the tuning for you, the Wizard points the user in the right direction by filling out the basic parameter values.
Used primarily in naturally aspirated applications, MAP/TPS Blend makes it possible for aggressively cammed engines to achieve a stable idle despite having an erratic MAP signal under low-load conditions. This is done by the referencing fuel delivery to throttle position sensor voltage at low engine speeds, slowing "blending" it out as the revs rise.
What's next?
In our next installment, we hope to pay our friends at UPRD a visit and roll our Project Impreza on to its four-wheel dynamometer for some real numbers. For now, we'll just revel in our newfound performance, knowing darn well that at only 70 percent injector duty cycle at stock fuel pressures, we still have a lot of room left to grow! Stay tuned. There's a lot more performance to come.
Sources
Electromotive Inc.
9131 Centreville Road
Manassas, Virginia 20110
(703) 331-0100 (TEL)
(703) 331-0161 (FAX)
www.electromotive-inc.com
RC Engineering
1728 Border Ave
Torrance, CA. 90501
(310) 320-2277 (TEL)
www.rceng.com
Vishnu Performance Systems
713 Liquidamber Place
Danville, CA 94506
(925) 648-9255
www.vishnuperformance.com